基于GEO数据库筛选慢性肾脏病相关差异表达基因及其初步验证

doi:10.3969/j.issn.1673-8640.2024.10.009

摘要/Abstract

摘要：

目的基于GEO数据库分析慢性肾脏病（CKD）患者的差异表达基因，并采用细胞学实验进行验证，以寻找CKD新的潜在靶点。方法从GEO数据库下载与CKD相关的微阵列数据（GSE66494数据集、GSE15072数据集和GSE30122数据集），采用生物信息学方法筛选出GSE66494数据集和GSE15072数据集中的差异表达基因，通过构建蛋白质-蛋白质互作（PPI）网络，确定Hub基因。在GSE30122数据集中分析筛选出的Hub基因的表达情况。采用NetworkAnalyst数据库构建微小RNA-Hub基因调控网络，筛选出与Hub基因关联居前5位的微小RNA。采用人近端肾小管上皮细胞系HK-2构建CKD细胞模型，根据加入的物质分为肿瘤坏死因子α（TNF-α）组（100 μg·L^-1 TNF-α）、TNF-α+infliximab组[10 ng·mL^-1 Infliximab（TNF-α特异性抑制剂）+100 μg·L^-1 TNF-α]和对照组（含等比例二甲基亚砜的培养基）。检测各组Hub基因和微小RNA的表达。结果共筛选出差异表达基因233个，其中表达上调20个、表达下调213个；从中筛选出10个Hub基因（HSP90AA1、IGF1、REN、MMP2、ACTA2、PXN、COL3A1、CALCA、DMD、LAMA2）和5个微小RNA（hsa-miR-29b-3p、hsa-miR-421、hsa-miR-29a-3p、hsa-miR-335-5p和hsa-let-7b-5p）。GSE30122数据集的分析结果显示，Hub基因在肾小管组织样本中表达更为显著。与对照组比较，TNF-α组出现明显的细胞坏死表现，细胞生长密度降低；HSP90AA1、IGF1、COL3A1、CALCA和hsa-miR-29b-3p、hsa-miR-421、hsa-miR-29a-3p相对表达量均显著升高（P<0.05）。与TNF-α组比较，TNF-α+infliximab组的细胞坏死表现明显改善，细胞生长密度增加；HSP90AA1、COL3A1和hsa-miR-29b-3p、hsa-miR-421、hsa-miR-29a-3p相对表达量均显著降低（P<0.05）。结论基于GEO数据库筛选出10个与CKD相关的Hub基因和5个微小RNA，其中HSP90AA1、IGF1、COL3A1、CALCA、hsa-miR-29a-3p、hsa-miR-29b-3p、hsa-miR-421可能参与了CKD的发生、发展，或可作为CKD潜在的诊疗靶点。

关键词: 差异表达基因, 微小RNA, 慢性肾脏病

Abstract:

Objective Based on the GEO database，to analyze and verify the differentially expressed genes in patients with chronic kidney disease（CKD） by cytological experiments in order to find new potential targets for CKD. Methods The microarray data related to CKD（GSE66494 dataset，GSE15072 dataset and GSE30122 dataset） were downloaded from GEO database. The differentially expressed genes in GSE66494 dataset and GSE15072 dataset were screened by bioinformatics methods. Hub genes were identified by constructing protein-protein interaction（PPI） network. The expression of selected hub genes was analyzed in the GSE30122 dataset. The NetworkAnalyst database was used to construct a microRNA-hub gene regulatory network，and the top 5 miRNA with high association with hub genes were screened. The proximal telomeric renal tubular epithelial cell line HK-2 was used to construct a CKD cell model. According to the different substances added，they were classified into tumor necrosis factor-alpha（TNF-α） group（100 μg·L^-1 TNF-α），TNF-α+infliximab group [10 ng·mL^-1 infliximab（TNF-α specific inhibitor）+100 μg·L^-1 TNF-α] and control group（equal proportion of solvent dimethylsulfoxide）. The expression of hub gene and microRNA in each group was determined. Results A total of 233 differentially expressed genes were screened，which included 20 up-regulated genes and 213 down-regulated genes. Totally，10 hub genes（HSP90AA1，IGF1，REN，MMP2，ACTA2，PXN，COL3A1，CALCA，DMD，LAMA2） and 5 microRNAs（hsa-miR-29b-3p，hsa-miR-421，hsa-miR-29a-3p，hsa-miR-335-5p and hsa-let-7b-5p） were screened out. The analysis results of the GSE30122 dataset showed that the hub gene was more significantly expressed in renal tubular tissue samples. Compared with control group，TNF-α group showed obvious cell necrosis and decreased cell growth density. The relative expression levels of HSP90AA1，IGF1，COL3A1，CALCA and hsa-miR-29b-3p，hsa-miR-421，hsa-miR-29a-3p were increased（P<0.05）. Compared with TNF-α group，the cell necrosis in TNF-α+infliximab group was improved，and the cell growth density was increased. The relative expression levels of HSP90AA1，COL3A1 and hsa-miR-29b-3p，hsa-miR-421，hsa-miR-29a-3p were decreased（P<0.05）. Conclusions There are 10 hub genes and 5 microRNA selected from GEO database. HSP90AA1，IGF1，COL3A1，CALCA and hsa-miR-29a-3p，hsa-miR-29b-3p，hsa-miR-421 may be involved in the process of CKD development，which can be potential targets for CKD.

Key words: Differentially expressed gene, MicroRNA, Chronic kidney disease

中图分类号:

R446.7

张世球, 吕婕, 俞容, 王川玲, 林子艳, 朱永俊. 基于GEO数据库筛选慢性肾脏病相关差异表达基因及其初步验证[J]. 检验医学, 2024, 39(10): 975-984.

ZHANG Shiqiu, LÜ Jie, YU Rong, WANG Chuanling, LIN Ziyan, ZHU Yongjun. Screening and preliminary verification of differentially expressed genes related to chronic kidney disease based on GEO database[J]. Laboratory Medicine, 2024, 39(10): 975-984.

图/表 12

表1 引物序列

基因名称	上游引物（5'~3'）	下游引物（5'~3'）
HSP90AA1	GGTACCCTACGGGGAGCG	GTTAACAGGTGCCCTGCTTCTCA
IGF1	AGAGCCTGCGCAATGGAATA	TGGTGTGCATCTTCACCTTCA
REN	CAAGTGCAGCCGTCTCTACA	TGATTCCACCCACGGTGATG
MMP2	GGACTTAGACCGCTTGGCTT	GTGTTCAGGTATTGCATGTGCT
ACTA2	GCCTGAGGGAAGGTCCTAAC	CTTCACAGGATTCCCGTCTT
PXN	TCCCTGCCATCACTGTGAAC	CAGGCCCTGGATCTTGAAAT
COL3A1	TAAAGGCGAAATGGGTCCCG	GGCACCATTCTTACCAGGCT
CALCA	AGATCCTGGCTCAGAGAGGT	GCACATAGTCCTGCACCAGT
DMD	ACAGCTGGCATGGAAGATGAA	TTCGAGGAGGTCTAGGAGGC
LAMA2	CTTGAGTATGAAAGCAAGGCCAG	AGAGCTCCACAAAACCAGGC
β-actin	ACAGAGCCTCGCCTTTGCC	GATATCATCATCCATGGTGAGCTGG
hsa-let-7b-5p	GCGCGTGAGGTAGTAGGTTGT	AGTGCAGGGTCCGAGGTATT
hsa-miR-29a-3p	CGCGTAGCACCATCTGAAAT	AGTGCAGGGTCCGAGGTATT
hsa-miR-29b-3p	CGCGTAGCACCATTTGAAATC	AGTGCAGGGTCCGAGGTATT
基因名称	上游引物（5'~3'）	下游引物（5'~3'）
hsa-miR-335-5p	CGCGTCAAGAGCAATAACGAA	AGTGCAGGGTCCGAGGTATT
hsa-miR-421	GCGCGATCAACAGACATTAATT	AGTGCAGGGTCCGAGGTATT
U6	CAAATTCGTGAAGCGTTCCA	AGTGCAGGGTCCGAGGTATT

图1 GSE66494数据集和GSE15072数据集的火山图和差异表达基因的韦恩图注：（a）GSE66494数据集的火山图；表达上调（610个）、表达下调（1 350个）、表达无差异；（b）GSE15072数据集的火山图；表达上调（571个）、表达下调（2 338个）、表达无差异；（c）GSE66494数据集与GSE15072数据集上调基因的韦恩图；（d）GSE66494数据集与GSE15072数据集下调基因的韦恩图；（e）上调基因与下调基因的交互韦恩图。

图2 GO和KEGG富集分析注：（a）表达上调基因的GO富集分析；（b）表达下调基因的GO富集分析；（c）表达上调和下调基因的KEGG通路富集分析。

图3 差异表达基因的PPI网络和Hub基因筛选注：（a）差异表达基因的PPI网络；表达上调、表达下调；（b）Hub基因，颜色由红到黄表示连接度由高到低。

表2 基于degree算法获得的Hub基因

排序	基因	编码蛋白	评分/分
1	HSP90AA1	热休克蛋白90α家族A类成员1	16
1	IGF1	类胰岛素一号增长因子	16
3	REN	肾素	14
3	MMP2	基质金属蛋白酶-2	14
5	ACTA2	肌动蛋白α2，平滑肌	11
5	PXN	桩蛋白	11
5	COL3A1	Ⅲ型胶原蛋白 alpha 1 链	11
8	CALCA	降钙素相关多肽α	10
9	DMD	肌营养不良蛋白	9
9	LAMA2	层黏连蛋白亚基α 2	9

图4 肾小球和肾小管组织样本中标注Hub基因的差异表达基因火山图注：（a）肾小球组织样本中标注Hub基因差异表达基因的火山图；表达上调（0个）、表达下调（1个）、表达无差异（7个）；（b）肾小管组织样本中标注Hub基因差异表达基因火山图；表达上调（2个）、表达下调（0个）、表达无差异（5个）。

图5 肾小球和肾小管组织样本中Hub基因诊断CKD的ROC曲线注：（a）肾小球组织样本中Hub基因的ROC曲线； HSP90AA1、 LAMA2、 CALCA、 COL3A1、 PXN、 ACTA2、 MMP2、参考线；（b）肾小管组织样本中Hub基因的ROC曲线； LAMA2、 DMD、 COL3A、 IGF1、参考线。

图6 微小RNA-Hub基因调控网络注：红点为Hub基因；蓝色为微小RNA；橘红色为关联度较高的微小RNA；红色线条表示基因与hsa-miR-29a-3p相连接，粉色线条表示微小RNA与COL3A1相连接，绿色线条表示hsa-miR-29a-3p与COL3A1之间的连接。

表3 CKD中关键微小RNA

微小RNA名称	degree值	相互作用的基因	中介中心性
hsa-miR-29a-3p	4	COL3A1、IGF1、MMP2、LAMA2	2 463.16
hsa-miR-335-5p	3	COL3A1、MMP2、PXN	4 229.57
hsa-let-7b-5p	3	COL3A1、DMD、HSP90AA1	2 971.08
hsa-miR-29b-3p	3	COL3A1、MMP2、LAMA2	217.37
hsa-miR-421	2	HSP90AA1、IGF1	4 480.20

图7 光学显微镜下对照组、TNF-α组和TNF-α+infliximab组细胞形态变化（×200）

图8 对照组、TNF-α组和TNF-α+infliximab组10个Hub基因相对表达量比较

图9 对照组、TNF-α组和TNF-α+infliximab组5种微小RNA相对表达量比较

参考文献 24

[1]	ROMAGNANI P, REMUZZI G, GLASSOCK R, et al. Chronic kidney disease[J]. Nat Rev Dis Primers, 2017, 3:17088. DOI PMID
[2]	谢晖, 沈瀚. 慢性肾脏病患者凝血功能变化及高凝状态影响因素分析[J]. 检验医学, 2021, 36(5):500-503. DOI
[3]	KDIGO Conference Participants. Genetics in chronic kidney disease:conclusions from a kidney disease:improving global outcomes(KDIGO) controversies conference[J]. Kidney Int, 2022, 101(6):1126-1141.
[4]	朱永俊, 俞容, 李晓燕, 等. Ang Ⅱ联合TNF-α促使肾小管上皮HK-2细胞发生程序性坏死[J]. 中国病理生理杂志, 2021, 37(6):1084-1090.
[5]	MÉNDEZ-GARCÍA L A, TREJO-MILLÁN F, MARTÍNEZ-REYES C P, et al. Infliximab ameliorates tumor necrosis factor-alpha-induced insulin resistance by attenuating PTP1B activation in 3T3L1 adipocytes in vitro[J]. Scand J Immunol, 2018, 88(5):e12716.
[6]	ZHU Q, SCHERER P E. Immunologic and endocrine functions of adipose tissue:implications for kidney disease[J]. Nat Rev Nephrol, 2018, 14(2):105-120.
[7]	GRANGE C, BUSSOLATI B. Extracellular vesicles in kidney disease[J]. Nat Rev Nephrol, 2022, 18(8):499-513.
[8]	SHEN A R, ZHONG X, TANG T T, et al. Integrin,exosome and kidney disease[J]. Front Physiol, 2021, 11:627800.
[9]	FEITENG C, LEI C, DENG L, et al. Relaxin inhibits renal fibrosis and the epithelial-to-mesenchymal transition via the Wnt/β-catenin signaling pathway[J]. Ren Fail, 2022, 44(1):513-524. DOI PMID
[10]	ZHAO W M, WANG Z J, SHI R, et al. Analysis of the potential biological mechanisms of diosmin against renal fibrosis based on network pharmacology and molecular docking approach[J]. BMC Complement Med Ther, 2023, 23(1):157.
[11]	GUREVICH E, SEGEV Y, LANDAU D. Growth hormone and IGF1 actions in kidney development and function[J]. Cells, 2021, 10(12):3371.
[12]	RAMANATHAN G, ELUMALAI R, PERIYASAMY S, et al. Renin gene rs1464816 polymorphism contributes to chronic kidney disease progression in ADPKD[J]. J Biomed Sci, 2016, 23:1. DOI PMID
[13]	CHENG Z, ZHANG X, ZHANG Y, et al. Role of MMP-2 and CD147 in kidney fibrosis[J]. Open Life Sci, 2022, 17(1):1182-1190. DOI PMID
[14]	GERRITS T, BROUWER I J, DIJKSTRA K L, et al. Endoglin is an important mediator in the final common pathway of chronic kidney disease to end-stage renal disease[J]. Int J Mol Sci, 2022, 24(1):646.
[15]	WEI J, XU Z, YAN X. The role of the macrophage-to-myofibroblast transition in renal fibrosis[J]. Front Immunol, 2022, 13:934377.
[16]	NOH K, BACH D H, CHOI H J, et al. The hidden role of paxillin:localization to nucleus promotes tumor angiogenesis[J]. Oncogene, 2021, 40(2):384-395.
[17]	甄永占, 胡刚, 章广玲, 等. 赖氨大黄酸对快速老化小鼠SAMP 10肾组织COL1A1、COL3A1和COL4A1表达的影响[J]. 第三军医大学学报, 2013, 35(17):1805-1808.
[18]	郑君芙, 邢淑丽, 宋瑞芬, 等. 降钙素基因相关肽与肾脏疾病[J]. 辽宁中医学院学报, 2006, 8(3):127-128.
[19]	VIJAYAN P, HACK S, YAO T, et al. LAMA2 and LOXL4 are candidate FSGS genes[J]. BMC Nephrol, 2021, 22(1):320.
[20]	DALGAARD L T, SØRENSEN A E, HARDIKAR A A, et al. The microRNA-29 family:role in metabolism and metabolic disease[J]. Am J Physiol Cell Physiol, 2022, 323(2):C367-C377.
[21]	TROJANOWICZ B, IMDAHL T, ULRICH C, et al. Circulating miR-421 targeting leucocytic angiotensin converting enzyme 2 is elevated in patients with chronic kidney disease[J]. Nephron, 2019, 141(1):61-74.
[22]	ZHU Y, CUI H, LV J, et al. Angiotensin Ⅱ triggers RIPK3-MLKL-mediated necroptosis by activating the Fas/FasL signaling pathway in renal tubular cells[J]. PLoS One, 2020, 15(3):e0228385.
[23]	ZHU Y, CUI H, LV J, et al. AT1 and AT2 receptors modulate renal tubular cell necroptosis in angiotensin Ⅱ-infused renal injury mice[J]. Sci Rep, 2019, 9(1):19450.
[24]	LIN Z, CHEN A, CUI H, et al. Renal tubular epithelial cell necroptosis promotes tubulointerstitial fibrosis in patients with chronic kidney disease[J]. FASEB J, 2022, 36(12):e22625.